[
https://issues.apache.org/jira/browse/AMQ-6432?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15849851#comment-15849851
]
Gary Tully commented on AMQ-6432:
---------------------------------
I think this should only occur if there is no disk sync, and thinking more if
sync is not disabled via enableJournalDiskSync=false then any non sync journal
write could lead to this. However, the gc should never try an read the current
journal file so it may have to coincide with a journal flip.
Having ack compaction force a sync before trying to scan may help here.
Getting better info out of the logging will be a first step, as would having a
reproducer.
> Improve 'Failed to load next journal location: null' warning output
> -------------------------------------------------------------------
>
> Key: AMQ-6432
> URL: https://issues.apache.org/jira/browse/AMQ-6432
> Project: ActiveMQ
> Issue Type: Improvement
> Affects Versions: 5.14.0
> Reporter: Martin Lichtin
>
> Seeing
> {noformat}
> 2016-09-19 15:11:30,270 | WARN | ournal Checkpoint Worker | MessageDatabase
> |
> emq.store.kahadb.MessageDatabase 2104 | 102 -
> org.apache.activemq.activemq-osgi - 5.14.0 |
> Failed to load next journal location: null
> {noformat}
> it'd be great to improve the output in such a case (Journal Checkpoint Worker
> ).
> Why not show the exception stack (it seems weird to only show the stack when
> debug level is enabled).
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
